The Effect of Laser Scanning Strategies on Texture, Mechanical Properties, and Site-Speciﬁc Grain Orientation in Selective Laser Melted 316L SS

Materials and Design 193 (2020) 108852

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier.com/locate/matdes

The effect of laser scanning strategies on texture, mechanical properties, and site-speciﬁc grain orientation in selective laser melted 316L SS

Jithin James Marattukalam a,⁎, Dennis Karlsson b, Victor Pacheco b, Přemysl Beran c,e, Urban Wiklund d, Ulf Jansson b, Björgvin Hjörvarsson a, Martin Sahlberg b a Department of Physics, Materials Physics, Uppsala University, Box 530, SE-75121 Uppsala, Sweden b Department of Chemistry- Ångström Laboratory, Uppsala University, Box 523, 75120 Uppsala, Sweden c Department of Neutron Physics, Nuclear Physics Institute, ASCR, CZ -25068 Řež, Czech Republic d Department of Engineering Sciences, Applied Materials Science, Uppsala University, Box 534, 75121 Uppsala, Sweden e European Spallation Source ERIC, Box 176, 22100 Lund, Sweden

HIGHLIGHTS GRAPHICAL ABSTRACT

• Laser scanning strategies strongly inﬂu- ence the crystallographic texture in as- printed components. • A ⟨100⟩ single crystalline-like texture is obtained in the direction of laser writing. • Site-speciﬁc texture control is achieved by selectively switching laser scanning strategies.

article info abstract

Article history: Selective laser melting has been used to demonstrate the striking effect of laser scanning strategies on the crys- Received 13 December 2019 talline texture in 316L SS. The aligned crystal orientation along the tensile direction (Z-axis) could be varied using Received in revised form 1 June 2020 the scanning strategy. A strong 〈100〉 single crystalline-like texture is obtained in the direction of the laser scan Accepted 2 June 2020 and a 〈110〉 texture was observed in the build direction when using a bidirectional scan without rotation. Fiber Available online 03 June 2020 texture was observed along the tensile direction when the bi-directional laser scanning vectors were rotated

Keywords: by 67° (Rot-scan) for each layer. The study highlights a correlation between laser scanning strategies with Selective laser melting resulting textures, microstructure, and mechanical properties in as-printed bulk 316L SS. The hardness, Young's Texture modulus, and ultimate tensile strength were significantly influenced by the final microstructure, crystallographic Austenitic stainless steel texture, and porosity. Furthermore, the applied laser scanning strategies made it possible to tailor crystallo- Solidification microstructures graphic textures locally within the component. This was demonstrated by printing characters with a fiber texture, in a matrix with ⟨100⟩ texture parallel to the Z-axis. © 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Additive manufacturing (AM) or 3D printing has become an indis- ﬁ ⁎ Corresponding author. pensable tool to fabricate complex geometries, which are often dif cult E-mail address: [email protected] (J.J. Marattukalam). to manufacture by conventional techniques. The process offers cost

https://doi.org/10.1016/j.matdes.2020.108852 0264-1275/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 2 J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852 reduction, product and process redesign, lead time reduction, and flex- ibility to manufacture parts with desired material properties [1–5]. Ad- ditive manufacturing has evolved from a mere prototyping and tooling technique to a direct part manufacturing method over the past few years [6]. A wide variety of materials have been processed through AM, including titanium, stainless steel, aluminum, nickel-based superal- loys, etc. [7–10]. Selective laser melting (SLM) is an AM technique where a scanning laser is used to melt along a certain direction in the powder bed. Typi- cally, high cooling rates and high thermal gradients are achieved in the SLM process, which results in a highly anisotropic solidification of the melt pool [11]. The thermal gradient and solidification front velocity can be used to predict the microstructural growth and morphology, while the cooling rate determines the size of the dendrites formed. Pro- cess parameters such as laser power, scan speed, hatch spacing, scanning strategies, and even sample orientation in the build plate have been reported to significantly alter the crystallographic texture during the process [12–15]. As crystalline texture is known to affect, Fig. 1. A secondary electron image showing the morphology of gas atomized 316L SS powder. e.g., mechanical and corrosion properties of a component, its control can be used to manufacture parts with desired material properties [16]. Several authors have discussed the impact of laser scanning strategy on the obtained texture in Ni-based alloys. For example, unidirec- Table S1. The printing was carried out in an EOS M100 (EOS GmBH, fi tional laser scanning was found to result in a fiber texture, while Germany) system equipped with a Yb ber laser. Different parameters bidirectional scanning of the laser gave rise to rotated cube-like texture including laser power, scan speed, hatch spacing, and hatch overlap, in the printed components [17]. Furthermore, a switching from trans- were varied in order to produce dense and crack free samples. The sam- verse anisotropic to isotropic and vice-versa has been demonstrated in ples used to study the effect of different scanning strategies were proc- the SLM of IN738 [18]. Crystallographic texture control in nickel-base essed using the optimised parameters, with a laser power of 107 W, μ superalloy Inconel 718 was demonstrated by varying the energy density scan speed of 800 mm/s, and hatch spacing of 70 m, to obtain the μ input to the powder bed using different electron beam scan strategies to best microstructural features. The thickness was set to 20 m for each μ locally change the grain orientation at precise locations in the as-build powder layer deposition. The laser spot diameter was 40 m, and bi- fl component [19]. A strong influence of laser scanning strategy on texture directional laser scanning was used. The process chamber was ushed has also been observed when processing wide range of other metals and with argon to minimize the presence of impurities in the material. alloys [20–22]. Three different scanning strategies were used in the printed parts, as One of the most common alloys is 316L SS, on which influences of shown in Fig. 2, Z-scan (along the raking direction of the powder bed), scan strategy on texture have been reported by Wei et al. [23]. They Y-scan (perpendicular to the raking direction), and 67° rot-scan with bi- showed that orientation of the primary dendrites was at 60° for all de- directional laser scanning vectors. The 3D computer-aided design (CAD) posited layers with unidirectional scanning laser scan, and for bi- models for the tensile samples were generated using Magics software, directional laser scan, it was found to be 90° in the neighboring layers. from Materialise, which is then numerically sliced into a number of Furthermore, in a recent study by Andreau et al. the effect of gas flow layers across the cross-section using EOS RP Tools 1.6 software. These and direction of laser scanning on crystallographic texture in 316L SS sliced layers are then recreated sequentially on a carbon steel build was discussed [24]. They observed that the crystallographic texture ap- plate to fabricate the components using an EOS PRINT software. peared to be dependent on the relative direction of laser scanning and The samples were cut (XY plane) and the pieces were polished using fi gas flow, which was attributed to an absorbing effect of the metal a series of SiC emery papers, diamond paste, and nally, silicon dioxide fi vapor plume. Scanning in the same direction as the gas flow was to obtain a mirror nish. The samples were etched in diluted aqua regia found to affect the depth of the melt pool, due to local attenuation of for the 10 s to reveal the melt pool patterns, which were observed using the laser beam's effective energy density transmitted to the powder Olympus AX70 optical microscope. The relative densities of as-printed bed. samples were determined using the Archimedes principle. The weight The aim of this work is to demonstrate how laser scanning strategy of the sample (W1) was measured in air at room temperature (RT). can be used to control texture during SLM, and even obtaining single The samples were then immersed in water and the weight (W2) was crystals of 316L SS. The microstructure and crystallographic textures measured at RT. The relative density of the sample was calculated as fol- formed during the process are correlated to the mechanical properties. lows; The use of laser scanning strategies to control the final shape of the ðÞ part and to simultaneously imprint a predesigned spatially resolved ðÞ¼% W1 x density of water : ð Þ Relative density ðÞ− : 100 1 crystallographic grain orientation pattern during the SLM process is W1 W2 x798 also demonstrated. The density of water at RT was taken to be 0.9978 g/cm3. The theo- 2. Materials and methods retical density of 316L SS is taken as 7.98 g/cm3. The weight of the samples was measured using an electronic balance with 0.1 mg accuracy. 316L SS gas atomized powders are produced by EOS, GmBH using Electron backscatter diffraction (EBSD) mapping was carried out industrial-grade feedstock materials. The as-received powder particles using an acceleration voltage of 20 kV and a beam current of 10 nA at were sieved using a 63 μm meshed sieve to remove oversized particles. 14 mm working distance in a Zeiss Merlin scanning electron microscope The sieved powder consists mostly of spherical particles, as shown in (SEM). The samples were mounted at 70° using a pre-tilted sample Fig. 1. The powder size distribution for 316L SS was determined accord- holder, and the data was recorded in the AZtec HKL software utilizing ing to ASTM B822/ISO13320, and the D10, D50 and D90 values are a Nordlys Max detector. The pole-figures were reconstructed from the 20 μm, 40 μm, and 60 μm respectively. The chemical composition of EBSD maps using a half-width of 10° to analyze the grain orientation the used virgin 316L stainless steel powder is shown in supplementary with respect to scan strategies in the printed parts. X-ray diffraction J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852 3

Fig. 2. Schematic view of laser scan strategies used for the fabrication of 316L SS samples. (a) laser scan direction along the Z direction, (b) laser scan direction along Y direction, and (c) laser scan rotated by 67° for each layer. Laser scanning strategies are illustrated in the YZ plane. The direction of gas ﬂow is parallel to the Y-axis, indicated by blue arrows. The region marked in yellow illustrates the planes (XY) used for structural investigations.

was performed on the XY planes in the as-processed samples. The mea- samples, which were machined according to ASTM E8M standard di- surements were carried out using D8 Bruker Twin-Twin diffractometer, mension with a 20 mm gauge length and 4 mm diameter. The samples with Cu-Kα radiation operating at 40 kV and 40 mA. 2θ angle was varied were pulled along the tensile direction, as indicated by the Z-axis in from 30° to 100° with a scan rate of one degree per minute. Fig. 2. The deformation rate was 0.005 mm/s, corresponding to an ap- In order to determine the bulk texture of the as-printed components, proximate strain rate of 0.00016/s. To allow estimation of fracture elon- neutron diffraction was used, enabling measurements on a large sample gation, two lines at 5 mm apart were carefully inscribed on the samples, volume of 10 × 10 × 10 mm3. This ensures that observations are repre- and the distances were measured before and after the tests (see supple- sentative of the whole sample. Samples are fixed on the Euler goniom- mentary Fig. S2). eter and rotated along φ (0–360°) and χ (0–90°) angles with a step of A possibility of fabricating components with desired geometries and 5°. Variation of neutron diffracted intensities for 4 individual reflections predetermined spatial three-dimensional crystallographic texture by [(111), (200), (220), and (311)] were collected for each orientation on using specific laser scanning strategies was also attempted. To demon- the MEREDIT instrument at the Nuclear Physics Institute, Czech strate site-specific switching of crystallographic textures, a rectangular Republic, using a neutron wavelength of 1.46 Å. The data was analyzed block with dimensions, 20 × 10 × 10 mm3 was fabricated with the let- using JTEX software [25]. The orientation distribution function (ODF) ters AMUU throughout the entire thickness of the block. The Z-scan was calculated for each sample, and pole figures for predefined reflec- and Rot-scan strategies shown in Fig. 2(a, c) was used to melt the letters tions were calculated for easy comparison of all samples. and block, respectively. Vickers microhardness analysis was conducted on a CM series microhardness indenter using 300 gf (~2.94 N) load and a dwell time of 3. Results and discussions 15 s. The load was chosen to 300 gf to avoid visual perception problems when measuring the length of the indentation tips on the sample, and 3.1. Effect of scan strategies on microstructure and crystallographic texture possibly avoiding potential data scatter [26]. The hardness with standard deviation was determined in all samples. Measurements were per- X-ray diffraction performed on 316L SS samples with different scan formed in the tensile direction (Z-axis). Rectangular samples strategies in the XY plane show preferential crystallographic orienta- 3 60 × 10 × 10 mm in dimension (see supplementary Fig. S1) were tions, as seen from Fig. 3. A change in relative intensities was observed manufactured for the uniaxial tensile test. Each series consisted of 3 for the (200) and (220) peaks when measured in the XY plane (Fig. 3

Fig. 3. X-ray diffraction patterns of (a) 316L SS powder and in-plane texture measured perpendicular to the XY plane indicated by the yellow area when scanned (b) along the Z-axis (c) Y- axis, and (d) 67° rotation. 4 J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852

(b, c)), as compared to the gas atomized powder in Fig. 3(a). This indi- rotated by 67° irregular half ellipse melt pool patterns are observed. cated the presence of texture effects in the XY plane when the ‘Z’ and The grains are also observed to grow and extended beyond the melt ‘Y’ scan strategies were used. However, rotation of scanning vectors by pool boundaries along the build direction (X-axis), as shown in Fig. 5(c). 67° in the sample results in similar intensities as observed in the pow- The influence of subsequent thermal cycling during multilayer selec- der, indicating a random-like distribution of grains in the XY plane tive laser melting plays a significant role in microstructure formation. (Fig. 3(d)). An important control parameter that influences the morphology of the The solidification microstructural features obtained during SLM is solidification microstructure is the ratio between Thermal gradient well studied in literature and is similar to the microstructural aspects (TG) and solidification velocity (VS). The effect of this parameter (TG/ observed during the welding process [27]. The microstructural features VS) on multilayer SLM and its role in the formation of a columnar in the fusion zone are observed to have different morphologies such as grain structure is elucidated elsewhere [34]. planar, cellular, and dendritic structures, and these are governed by To highlight the observed changes in crystallographic texture and temperature gradient, grain growth rate, undercooling, and chemical melt pool morphologies with different scanning strategies, detailed mi- composition. With increasing constitutional undercooling, the crystalli- crostructural studies were performed using scanning electron micros- zation mode changes from planar to cellular and finally to dendritic so- copy (SEM) and electron backscattered diffraction (EBSD). The melt lidification mode. A planar growth pattern is observed along line OA and pool morphologies formed are significantly different with respect to a dendritic growth mode along OB, as illustrated in Fig. 4. The different each other and are dependent on the scan strategies used, as illustrated solidification modes can be attributed to the variation of the tempera- in Fig. 5. When viewed in the XY plane from Fig. 5(a, d), a continuous ep- ture distribution across OA and OB within the melt pool. A higher tem- itaxial growth of columnar grains is observed along the centerline of the perature gradient across OA favours planar growth; meanwhile, along melt pool (indicated by black arrows in Fig. 5(a)). The observed epitax- OB, the temperature gradient is smaller, which results in the formation ial growth is a dominant feature in microstructure evolution, and the of constitutional undercooling zone and induces dendritic growth mode formation of such feature is attributed to the direction of thermal gradi- (along OB). Similar observations on grain growth patterns have been ent being vertical along the centreline of the melt pool, as frequently re- previously reported by other researchers on selective laser melting of ported in the literature [16,35,36]. However, the growth of these vertical steels [28–30]. columnar grains is sometimes found to be broken and discontinuous Fig. 5 shows optical images highlighting the melt pool patterns along the melt pool centers. This is possibly due to a dominant lateral formed during the SLM process. Three types of melt pool patterns are migration of the solid-liquid interface, which disrupts their continuous clearly observed with respect to each scan strategy, as shown in Fig. 2. growth [16].The Marangoni flow, which causes the convective heat In Fig. 5(a), a nail-head melt pool pattern is observed in the XY plane transfer and fluid flow within the melt pool, could be one of the factors with respect to a Z-scan strategy. This mode of melt pool formation which can influence the heat flux direction due to its heat-mass transfer has been reported previously by Fabbro et al. on laser welding, and a effect and thereby affect the growth orientation of the dendrites [37,38]. similar observation was made by King et al. for powder bed fusion pro- The backward flow of the melt material is the primary heat transfer cess in 316 L stainless steel [31,32]. A possible reason for the formation path, which is also the main direction of solidification. It is observed that of such melt pools is a positive thermocapillary flow (Marangoni effect). a strong ⟨100⟩ crystallographic texture is always obtained along the di- This positive convective flow is directed from the center of the melt pool rection of the laser scan (see Fig. 6(a, d)), in agreement with previous towards the bottom, dissipating most of the heat downwards rather studies [30,37,39]. The 180° degree rotation of the bi-directional laser than towards the sides of the melt pool [33]. Columnar grains were scan strategy was not found to affect the crystallographic growth of found to extend vertically in these melt pool patterns. the ⟨100⟩ along the scan direction as multiple rotations of 90° will result Fig. 5(b) shows the melt pool pattern in the XY plane resulting from in the same crystallographic orientations for cubic symmetries [24]. It is a Y-scan strategy. Long bands are observed, which are generated due to known that in cubic crystal symmetry, ⟨100⟩ has 6 equivalent crystal ori- melt pool flow along the scan. Columnar grains are observed to grow entations. If the [001] crystallographic direction is fixed parallel to the perpendicular to these horizontal band like melt pool traces. This scan direction, the other equivalent crystallographic orientations could shows that columnar grains mainly grow on a plane perpendicular to rotate within the track about a fixed axis when subjected to a radial the scanning direction, as seen in Fig. 5(b). When the laser scan is thermal dissipation [37]. From Fig. 6(d), it is observed that a strong 〈110〉 texture is obtained along the build direction and a 〈100〉 texture along the scan direction when the scan is bidirectional without rotation. This is an important result obtained from this study, which shows that the growth of the columnar cells preferentially occurred in the 〈100〉 crystallographic direction within the plane perpendicular to the scan direction. A recent study by Pham et al. attributed the formation of such a crystallographic texture to the side-branching of columnar dendritic cells in the microstructure [36], as shown in Fig. 4. The underlying mechanism of this phenomenon is due to the misalignment between the existing columnar cells and the thermal gradient, which promotes side-branching due to the perturbations on the existing cells on the sides of the melt pool. The misalignment between the existing thermal gradient and growth direction of columnar cells is a result of a decrease in nucleation energy by conserving crystal orientation between two adjacent melt pools [22]. The side-branching columnar cells grew along a 〈100〉 crystallographic direction, inclined about 45° with respect to the build direction, as shown in Fig. 7(a). The orientation relationship of the crystals formed within the melt pool with respect to the scan direction (Z- scan) is shown schematically in Fig. 7 for better visualization of single crystalline-like texture formation during SLM. Fig. 4. SEM image showing planar and dendritic growth patterns along OA and OB, Fig. 5(b) shows the growth of the columnar grains perpendicular to respectively, which is dependent on the temperature gradient within the melt pool. the elongated band-like melt pool traces when scanned along the Y-axis J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852 5

Fig. 5. Optical (a–c) and SEM (d–e) images in the XY plane showing morphologies of melt pool patterns and grains formed during selective laser melting of 316L SS austenitic steel with speciﬁc scan strategies. (a, d) scan vectors along Z-axis (b, e) scan vectors along Y-axis, and (c, f) scan vectors rotated by 67°. The inset indicates the respective scanning strategies in the YZ plane for each sample. The black arrows in panel (a) indicate the growth direction of columnar lamellar microstructure.

with a bidirectional scan strategy. From Fig. 6 (b, e) it can be observed observed in Fig. 5(f) and 6(c). The formation of such disoriented melt that the 〈110〉 crystallographic direction perpendicular is to the XY pool traces in each layer disrupts the development of a single plane (along Z-axis, tensile direction). Fig. 5(f) shows the melt pool crystalline-like texture with the rotation scan strategy. Pole figures traces when the scanning vectors are rotated. In the 67° Rot-scan strat- from Fig. 6(f) indicate a strong 〈110〉 texture perpendicular to the YZ egy, the melt pool centers of the neighboring layers do not coincide with plane, which is along the direction of the build (X). The diffused rings each other, and the melt pools appear as half ellipses when viewed in seen from the {110} pole figure demonstrate a fiber texture along Z the XY plane. They exhibit different morphology in each layer, as (tensile direction), as illustrated in Fig. 6(f).

Fig. 6. EBSD maps (XY plane) and pole ﬁgures in (XY and YZ plane), showing crystallographic grain orientations during selective laser melting of 316L SS austenitic steel when scanned (a, d) along Z-axis, (b, e) Y-axis and (c, f) at 67° rotation. The inset indicates the respective scanning strategies in the YZ plane for each sample. 6 J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852

Fig. 7. (a) Enlarged view of an EBSD image from the black square in Fig. 6(a) with a chevron-like pattern showing the growth of grains along the 〈100〉 crystallographic direction. (b) Schematic illustration showing the melt pool pattern formed during SLM with respect to the bi-directional scanning strategy along the Z-axis in the sample. (c) Crystallographic orientations of columnar cells formed within the melt pool during a bidirectional scan without rotation. The build direction (BD) is along the X-axis, and scanning direction (SD) is along the Z- axis, as illustrated above. Region 1 illustrates the continuous epitaxial growth of columnar grains, and region 2 illustrates the side-branching of dendrites growing along 〈100〉 crystallographic direction. For further details, please see the text.

Neutron diffraction was used to determine the bulk texture in the The effect of gas flow on microstructure and the crystallographic tex- samples fabricated using the different scanning strategies. It is observed ture was also investigated in the present study. It was concluded from that the {110} planes are oriented perpendicular to the build direction the neutron diffraction and EBSD studies that a strong 〈100〉 single (X-axis), while the {100} crystallographic planes are perpendicular to crystalline-like texture always evolves along with the motion of bi- the laser scan direction, as shown in Fig. 8 (a, b), which is also in agree- directional laser irrespective of the direction of the scanning (Z scan or ment with the EBSD data. Typically, the alignment of the 〈100〉 crystal- Y scan). The melt pool size and shape along the direction of the laser lographic direction along the scan direction is explained by the thermal scan was also qualitatively observed to understand the effect of gas gradient induced during SLM [22,23,40]. The rotation of the scan vectors flow on as-processed 316L SS, processed with Z and Y scanning strate- in every layer by 67° results in a more complex and varied melt pool gies (see supplementary Fig. S4). It was concluded from the present traces. This scanning strategy results in the fragmentation of the colum- study that there was no detectable effect of vapor plume on the crystal- nar grain structure (see Fig. 6(c)). The specimen built with the rotation lographic texture with respect to the direction of the gas flow, in stark scan strategy exhibits a fiber texture with a strong alignment of 〈110〉 contrast to the conclusions made by Andreau et al. [24]. along the build direction (X-axis) and a random distribution of 〈100〉, To quantify the size of cellular subgrains, the intragranular cell spac- 〈110〉 and 〈111〉 crystallographic directions along the tensile direction ing [41] was evaluated, as illustrated in Fig. 9. The intragranular cell (Z-axis) as illustrated in Fig. 6fandFig. 8c. The ideal texture components spacing value can be calculated by using the area method by the follow- for the as-processed samples are illustrated in supplementary Fig. S3. ing equation [29,42]:

Fig. 8. Calculated neutron pole ﬁgures showing preferential crystallographic grain orientations, when the scanning vectors are along (a) Z-axis, (b) Y-axis, and (c) rotated by 67° respectively. A strong 〈100〉 texture is obtained along the laser scan direction, and 〈110〉 texture is observed along the build direction (X-axis). J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852 7

Fig. 9. SEM micrographs in the X-Y plane (a) Microstructure at different length scales, which includes melt pool fusion boundaries, grain boundaries, columnar substructure, and cellular structure. High magniﬁcation SEM images of the cellular sub-grain structure formed during (b) Z-scan (c) Y-scan, and (d) Rot-scan strategies. The inset indicates the respective scanning strategies in the YZ plane for each sample. The white rectangle in Fig. 9(b), indicates the target zone for calculating the intragranular cell spacing.

[47,48]. The typical size of these cellular structures is b1 μm, with a 1 = Intragranular cell spacing ¼ ðÞAjN 1 2 ð2Þ cell wall thickness of the order of a few hundred nanometres [44]. M Owing to their small size, they can provide more barriers for dislocation motion in selectively laser melted austenitic stainless steels than in con- Where ‘M,’ is the magnification factor of the SEM micrograph, ‘N’ is ventional stainless steels that lack microstructural features at this scale. the number of cellular subgrains in the target zone, as shown in Fig. 9 Experimentally observed values of Young's modulus in as-processed (b), indicated by the rectangle and ‘A’ is the area of the rectangle. The samples, along with scanning strategies and resulting crystallographic fi average value of intragranular cell spacing with standard deviation textures, are shown in Table 1. The results show a signi cant effect of was determined for each scanning strategy by considering different the applied scan strategies on Young's modulus in the samples, mea- areas in the SEM images with similar magnification having a cellular sured along the Z-axis. The Young's modulus was observed to change structure. The intragranular cell spacing showed a decreasing trend with scanning strategy from 129 ± 3 GPa (Z-scan) to 193 ± 16 GPa when the scanning strategies changed from Z-scan (〈100〉 single (67°Rot-scan), as illustrated in Table 1. Since Young's modulus depends crystalline-like texture) to Y-scan (〈110〉 single crystalline-like texture) on the strength and density of interatomic forces, its value is sensitive to fi as shown in Table 1. The change in intragranular cell spacing may be due the atomic arrangement along a speci c crystallographic direction [58]. to a variation of thermal history and degree of undercooling in the melt In general, the elastic modulus varies with interatomic distances in a pool with different scan strategies. single crystal, and for an FCC system, it is ordered accordingly as E [111] N E [110] N E [100] [59,60]. Since the texture along the tensile direction 3.2. Effect of scan strategies on mechanical properties (Z-axis) differs with respect to scanning strategies, the observed variation in Young's modulus is expected, as shown in Table 1. Microhardness measurements showed marginal changes in hard- Fig. 11 shows the combined stress-strain curve for samples printed ness in the tensile direction (Z-axis), as shown in Fig. 10. However, no with different laser scanning strategies. The yield strength was found significant differences in hardness were observed between the build to increase from 554 ± 5 MPa in samples with [100] texture to 603 ± fi and the tensile directions in the as-processed samples despite the differ- 2 MPa in samples with ber texture, as shown in Table 1. The change ence in microstructure and crystallographic grain orientations. The mar- in tensile yield strength could be due to several factors such as micro- ginal change in observed hardness could be due to the small variations structure (intragranular cell spacing), crystallographic texture (Taylor fl in intragranular cell spacing, illustrated in Table 1,from0.40μmto factor), and porosity in as-printed samples. The in uence of the 0.76 μm. A Hall-Petch type relation can be fitted between hardness above-mentioned factors on the mechanical properties is discussed in and intragranular cell spacing, as illustrated in Fig. 10. Cellular subgrain the following section. structures formed during laser powder bed fusion (LPBF) act as barriers The mechanical properties of the samples appear to be strongly in- fl for dislocations and induce high densities of entangled dislocations uenced by the presence of intragranular cellular structures. The ob- [43–46]. These cellular networks formed in 316L SS during LPBF are served cellular structures formed, see Fig. 9(b, c, d) result from high found to be stabilized by segregated alloying atoms and minor mis- cooling rates and non-equilibrium conditions in the SLM process [49]. orientations developed during rapid solidification of the melt pool The morphology of the cellular structure and size of the cells are 8 J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852

Table 1 Tensile properties and intergranular cell spacing of AM 316L SS with speciﬁc in-plane textures formed with different laser scanning strategies, along the Z-axis.

Scan strategy Resulting texture Young's modulus Yield strength Ultimate tensile strength Elongation at fracture Intragranular cell spacing (μm) (GPa) (MPa) (MPa) (%)

Z-scan (100) 129 ± 3 554 ± 5 649 ± 3 26 ± 1 0.76 ± 0.07 Y-scan (110) 167 ± 3 555 ± 11 613 ± 5 42 ± 6 0.53 ± 0.03 Rot-scan Fiber 193 ± 16 603 ± 2 690 ± 2 32 ± 2 0.40 ± 0.03

influenced by laser scanning strategies, and more specifically, by the strategies along the loading direction (Z-axis) is evaluated from EBSD resulting temperature gradients during the laser melting process. (see supplementary Fig. S5). The Taylor factor distribution is mainly be- The cellular structure exists in the columnar grains and is present as tween 2.3 and 2.6 for 〈100〉 textured sample (Z-scan strategy) along the an intragranular feature. Hence, a cellular structure cannot be loading direction (Z-axis). For the fiber textured sample (Rot-scan), the interpreted as a grain structure because the adjacent cells have similar distribution of the Taylor factor became broader and has an average crystallographic orientation, as previously reported elsewhere [49]. Al- value of 3, along the loading direction (Z-axis). Meanwhile, an increase though cellular structures cannot be considered as grains, they play a in the fraction of grains with the Taylor factor over 3.4 is observed for significant role in acting as barriers to dislocation motion during defor- 〈110〉 textured sample (Y-scan strategy) along the loading direction mation. A Hall-Petch type correlation between tensile strength and (Z-axis). The Taylor factor is minimum for the 〈100〉 crystallographic intragranular cell spacing have been previously reported in [50,51]. texture in a cubic material (see supplementary Fig. S5(a)), this can Both the yield strength and ultimate tensile strength showed a linear partly explain the lower yield strength observed in the as-printed sam- correlation with the reciprocal of the square root of intragranular cellu- ple with 〈100〉 crystallographic texture [56,57]. Meanwhile, a marginal lar spacing, indicating the importance of cellular structures in the increase in ultimate tensile strength for the sample with 〈100〉 texture, strengthening of selectively laser melted 316L SS components. when pulled along the Z-axis, could be due to the formation of the co- The size of these cellular structures formed during SLM with differ- lumnar lamellar microstructure (CLM) as observed and is indicated by ent scanning strategies is determined using the ‘area method’ as de- black arrows in Fig. 5(a) and Fig. 7(a). The formation of such microstruc- scribed above. The mean cellular size is found to vary from 0.76 μmto tural features was found for the first time by Sun et al. during the selec- 0.40 μm, illustrated in Table 1. However, a Hall-Petch type correlation tive laser melting of austenitic stainless steel [16]. A similar between the yield and ultimate tensile strength with intragranular cell microstructural feature with CLM layers is observed in the sample spacing was not observed in the present study. This could be due to with the Z-scan strategy, as presented in our study. The CLM layers ob- the influence of the two other factors (Taylor factor and porosity) as served showed a discontinuous and broken pattern along the direction discussed below; of the build. These microstructural features can act as boundaries be- The variation in yield strength can be related to variations in the Tay- tween the textured regions and possibly hinder dislocations move- lor factor, ‘M’ [52], which is related to the flow stress σ by: ments, leading to an increased, ultimate tensile strength. A third factor influencing the yield and ultimate tensile strength is = σ ¼ MαbGðÞρ 1 2 ð3Þ the difference in porosity [58–60]. The change in porosity and tensile strength for sample made with different scan strategies (top sectional view) during the process are shown in Fig. 12. The relative density where ‘b’ is the absolute value of Burgers vector, ‘G’ is the shear mod- and average porosity in as-processed bulk samples determined using ulus, ρ is the total dislocation density, and α is a constant of the order of the Archimedes principle, as shown in Table S2. A high densification 0.2–0.35 [53,54]. Since the Taylor factor is an orientation factor which level greater (≥99.5%) was observed in samples. The pores were not depends on the texture of the materials and on the crystallographic na- evenly distributed throughout the sample and were found to be concen- ture of the assumed slip systems, materials with higher Taylor factors trated more in certain areas as compared to the rest of the sample [61]. require greater axial stress to be plastically deformed [55]. The Taylor factor distribution for as-processed samples with different scanning

Fig. 11. Stress vs. strain in as-printed 316L SS with different scanning strategies. The Fig. 10. Hardness as a function of the intragranular cell spacing, measured in the tensile direction of the applied load is along the Z-axis, which is illustrated by the green arrows direction (Z-axis). in the insets. J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852 9

occurs initially by slip, but at higher strain, deformation occurs by twinning [70]. Grains with 〈110〉 and 〈111〉 initially oriented along the tensile direction (Z-axis) are especially prone to mechanical twinning [71]. This effect will promote strain hardening of the material with the 〈110〉 texture and delay the onset of necking, resulting in the excellent elongation during deformation (see supplementary Fig. S7). The mechanical properties with respect to microstructural and texture evolution in deformation twinned austenitic stainless steel has been discussed in detail elsewhere [72,73].

3.3. Site-speciﬁc crystallographic grain orientation by use of laser scanning strategies

Finally, as a part of this study, the site-specific crystallographic orientation of grains within a given component was demonstrated by design- ing a 3D CAD model with box and letters, as shown in Fig. 13(a). From the above discussions, it is observed that strong single crystalline-like and fiber textures could be generated along the tensile direction (Z- axis) by switching the scanning strategies. This knowledge was used to create different textures within the component by assigning a Z- Fig. 12. Variation in porosity, ultimate tensile strength, and percentage elongation with scan and Rot-scan strategy to the box and letters, respectively. The different laser scanning strategies during the SLM process, measured along the Z-axis. box shows a preferred 〈100〉 orientation in the tensile direction, while a fiber texture is observed inside the letters. This is in agreement with From optical microscopy, the highest area fraction of porosity was ob- the observations shown in Fig. 6(a, c) and further implicates that the served in the Y-scan sample (see supplementary Fig. S6). This increase texture is changeable within a single component. Additionally, the mi- in the amount of porosity during SLM can be due to the phenomenon crostructure and melt pool morphologies formed in the component of spatter during SLM [62–64]. The amount of spatter generated with re- can be traced back to the initial observations shown in Fig. 5(a, c) and spect to laser scanning strategies has been previously reported else- Fig. 6(a, c). Microstructure and texture control in different parts of the where [65–67]. Spatter results in detrimental effects on the component opens up possibilities to achieve tunable properties de- components and often results in higher porosity and, consequently, a pending on specific needs for the application [19]. reduced yield strength. The observed variation in density due to the difference in scanning strategy could partly explain the lower yield and 4. 5. Conclusions tensile strength in the 〈110〉 single crystalline-like textured sample compared to the 〈100〉 textured sample as shown in Fig. 12. The present study shows that a suitable combination of laser power, The porosity observed in the laser melted samples could also be due scan speed, and scanning strategy can be used to determine the to the phenomenon of balling (resulting from the spatter), which was resulting texture in the SLM processed 316L SS. The growth of columnar observed at the laser melted surface in the present study. The presence grains within the plane perpendicular to the scanning direction was of these micrometer scaled balls on the surface of the laser melted sam- found to be essential for the development of a strong single ples can lead to improper powder bed spreading and affect the final part crystalline-like texture. A strong 〈100〉 and 〈110〉 single crystalline-like density. The formation of such features could be ascribed to laser- texture was observed in the direction of laser scan and build direction, induced melt splashes caused by capillary instability of the melt and is respectively, in samples printed with a bidirectional scanning strategy correlated to laser scanning speeds, as previously investigated by Gu without rotation. The rotation of scans by 67° resulted in a mismatch et al. [68]. Porosity in the as-processed parts can also arise due to the of the positions of melt pool in each layer, which broke the epitaxy of co- presence of entrapped gases, resulting from oxide decomposition at lumnar growth and resulted in fiber texture. Furthermore, it was shown high temperatures during the SLM process [69]. that the applied laser scan strategies greatly influenced the mechanical An increase in fracture elongation from 26% to 42% was also ob- properties. The hardness, Young's modulus, yield strength, and elonga- served with a change in scanning strategy. A study by Sinha et al. in tion at fracture were influenced by the final microstructure, porosity, 316L austenitic stainless steel demonstrated that tensile deformation and texture formed in these samples. This shows the possibility to

Fig. 13. (a) Computer-aided design file with alphabets AMUU and laser scanning strategies Z-scan and 67° Rot-scan used to create site-specific crystallographic grain orientations. (b) EBSD image of SLM component from 13(a) showing a single crystalline-like texture in the rectangular box and fiber texture within the letters AM. 10 J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852 manufacture components with both isotropic and anisotropic proper- [9] F. Abe, K. Osakada, M. Shiomi, K. Uematsu, M. Matsumoto, The manufacturing of fi hard tools from metallic powders by selective laser melting, J. Mater. Process. ties during a single SLM build. The site-speci c control of crystallo- Technol. 111 (2001) 210–213, https://doi.org/10.1016/S0924-0136(01)00522-2. graphic grain orientations within a given metal component was [10] W. Xu, M. Brandt, S. Sun, J. Elambasseril, Q. Liu, K. Latham, K. Xia, M. Qian, Additive demonstrated by selectively switching laser scanning strategies. manufacturing of strong and ductile Ti–6Al–4V by selective laser melting via in situ martensite decomposition, Acta Mater. 85 (2015) 74–84, https://doi.org/10.1016/j. actamat.2014.11.028. Data availability [11] V. Manvatkar, A. De, T. DebRoy, Heat transfer and material flow during laser assisted multi-layer additive manufacturing, J. Appl. Phys. 116 (2014), 124905. https://doi. The raw/processed data required to reproduce these findings cannot org/10.1063/1.4896751. [12] T. Ishimoto, K. Hagihara, K. Hisamoto, S.-H. Sun, T. Nakano, Crystallographic texture be shared at this time as the data also forms part of an ongoing study. control of beta-type Ti–15Mo–5Zr–3Al alloy by selective laser melting for the development of novel implants with a biocompatible low Young’s modulus, Scr. Mater. CRediT authorship contribution statement 132 (2017) 34–38, https://doi.org/10.1016/j.scriptamat.2016.12.038. [13] K. Kunze, T. Etter, J. Grässlin, V. Shklover, Texture, anisotropy in microstructure and mechanical properties of IN738LC alloy processed by selective laser melting (SLM), Jithin James Marattukalam: Conceptualization, Investigation, For- Mater. Sci. Eng. A 620 (2015) 213–222, https://doi.org/10.1016/j.msea.2014.10.003. mal analysis, Writing - original draft, Writing - review & editing. Dennis [14] A.A. Deev, P.A. Kuznetcov, S.N. Petrov, Anisotropy of mechanical properties and its correlation with the structure of the stainless steel 316L produced by the SLM Karlsson: Investigation, Formal analysis, Writing - original draft, Writ- method, Phys. Procedia 83 (2016) 789–796, https://doi.org/10.1016/j.phpro.2016. ing - review & editing. Victor Pacheco: Investigation, Formal analysis. 08.081. Přemysl Beran: Investigation, Formal analysis. Urban Wiklund: Super- [15] D. Gu, H. Chen, Selective laser melting of high strength and toughness stainless steel vision. Ulf Jansson: Supervision. Björgvin Hjörvarsson: Supervision. parts: the roles of laser hatch style and part placement strategy, Mater. Sci. Eng. A 725 (2018) 419–427, https://doi.org/10.1016/j.msea.2018.04.046. Martin Sahlberg: Conceptualization, Supervision. [16] S.-H. Sun, T. Ishimoto, K. Hagihara, Y. Tsutsumi, T. Hanawa, T. Nakano, Excellent mechanical and corrosion properties of austenitic stainless steel with a unique crystallographic lamellar microstructure via selective laser melting, Scr. Mater. 159 (2019) 89–93, https://doi.org/10.1016/j.scriptamat.2018.09.017. Declaration of competing interest [17] G.P. Dinda, A.K. Dasgupta, J. Mazumder, Texture control during laser deposition of nickel-based superalloy, Scr. Mater. 67 (2012) 503–506, https://doi.org/10.1016/j. The authors declare that they have no known competing financial scriptamat.2012.06.014. fl [18] F. Geiger, K. Kunze, T. Etter, Tailoring the texture of IN738LC processed by selective interests or personal relationships that could have appeared to in u- laser melting (SLM) by specific scanning strategies, Mater. Sci. Eng. A 661 (2016) ence the work reported in this paper. 240–246, https://doi.org/10.1016/j.msea.2016.03.036. [19] R.R. Dehoff, M.M. Kirka, W.J. Sames, H. Bilheux, A.S. Tremsin, L.E. Lowe, S.S. Babu, Site specific control of crystallographic grain orientation through electron beam additive Acknowledgments manufacturing, Mater. Sci. Technol. 31 (2014) 931–938, https://doi.org/10.1179/ 1743284714y.0000000734. This research is funded by the Swedish Foundation for Strategic Re- [20] L. Thijs, K. Kempen, J.-P. Kruth, J. Van Humbeeck, Fine-structured aluminium prod- search (SSF) within the Swedish national graduate school in neutron ucts with controllable texture by selective laser melting of pre-alloyed AlSi10Mg powder, Acta Mater. 61 (2013) 1809–1819, https://doi.org/10.1016/j.actamat. scattering (SwedNess) and by the Swedish Foundation for Strategic Re- 2012.11.052. search (SSF) through the project “Development of Process and Materials [21] L. Thijs, M.L. Montero Sistiaga, R. Wauthle, Q. Xie, J.-P. Kruth, J. Van Humbeeck, in Additive manufacturing”, Reference number GMT14-0048. Neutron Strong morphological and crystallographic texture and resulting yield strength anisotropy in selective laser melted tantalum, Acta Mater. 61 (2013) 4657–4668, measurements were performed at the CANAM infrastructure of the https://doi.org/10.1016/j.actamat.2013.04.036. NPI CAS Rez and the reactor LVR15 supported through MŠMT project [22] S.H. Sun, K. Hagihara, T. Nakano, Effect of scanning strategy on texture formation in No. LM2015056 and LM2015074. We are also grateful for all the valu- Ni-25 at.%Mo alloys fabricated by selective laser melting, Mater. Des. 140 (2018) 307–316, https://doi.org/10.1016/j.matdes.2017.11.060. able technical support by L. Edlén, N. Johansson, and E. Lindquist. [23] H.L. Wei, J. Mazumder, T. DebRoy, Evolution of solidification texture during additive manufacturing, Sci. Rep. 5 (2015) 1–7, https://doi.org/10.1038/srep16446. Appendix A. Supplementary data [24] O. Andreau, I. Koutiri, P. Peyre, J.-D. Penot, N. Saintier, E. Pessard, T. De Terris, C. Dupuy, T. Baudin, Texture control of 316L parts by modulation of the melt pool morphology in selective laser melting, J. Mater. Process. Technol. 264 (2019) 21–31, Supplementary data to this article can be found online at https://doi. https://doi.org/10.1016/j.jmatprotec.2018.08.049. org/10.1016/j.matdes.2020.108852. [25] B. Beausir, J.J. Fundenberger, Analysis Tools for Electron and X-ray Diffraction, ATEX- Software, Univ. Lorraine-Metz, 2017. [26] F. George, V. Voort, R. Fowler, Low-load Vickers microindentation, Adv. Mater. Pro- References cess. 170 (4) (2012) 28–33. [27] J.P. Oliveira, T.G. Santos, R.M. Miranda, Revisiting fundamental welding concepts to [1] E. Liverani, S. Toschi, L. Ceschini, A. Fortunato, Effect of selective laser melting (SLM) improve additive manufacturing: from theory to practice, Prog. Mater. Sci. (2019) process parameters on microstructure and mechanical properties of 316L austenitic 100590. – stainless steel, J. Mater. Process. Technol. 249 (2017) 255 263, https://doi.org/10. [28] K. Kempen, E. Yasa, L. Thijs, J.-P. Kruth, J. Van Humbeeck, Microstructure and me- 1016/j.jmatprotec.2017.05.042. chanical properties of selective laser melted 18Ni-300 steel, Phys. Procedia 12 [2] M.M. Kirka, Y. Lee, D.A. Greeley, A. Okello, M.J. Goin, M.T. Pearce, R.R. Dehoff, Strat- (2011) 255–263. egy for texture management in metals additive manufacturing, JOM 69 (2017) [29] D. Wang, C. Song, Y. Yang, Y. Bai, Investigation of crystal growth mechanism during 523–531, https://doi.org/10.1007/s11837-017-2264-3. selective laser melting and mechanical property characterization of 316L stainless [3] M. Rombouts, J.P. Kruth, L. Froyen, P. Mercelis, Fundamentals of selective laser melt- steel parts, Mater. Des. 100 (2016) 291–299. ing of alloyed steel powders, CIRP Ann. 55 (2006) 187–192, https://doi.org/10.1016/ [30] I. Yadroitsev, P. Krakhmalev, I. Yadroitsava, S. Johansson, I. Smurov, Energy input ef- S0007-8506(07)60395-3. fect on morphology and microstructure of selective laser melting single track from [4] M. Badrossamay, T.H.C. Childs, Further studies in selective laser melting of stainless metallic powder, J. Mater. Process. Technol. 213 (2013) 606–613, https://doi.org/ and tool steel powders, Int. J. Mach. Tools Manuf. 47 (2007) 779–784, https://doi. 10.1016/j.jmatprotec.2012.11.014. org/10.1016/j.ijmachtools.2006.09.013. [31] R. Fabbro, Dynamic approach of the keyhole and melt pool behavior for deep pene- [5] J.P. Kruth, L. Froyen, J. Van Vaerenbergh, P. Mercelis, M. Rombouts, B. Lauwers, Selec- tration Nd-Yag laser welding, AIP Conf. Proc. 1047 (2008) 18–24, https://doi.org/10. tive laser melting of iron-based powder, J. Mater. Process. Technol. 149 (2004) 1063/1.2999933. 616–622, https://doi.org/10.1016/j.jmatprotec.2003.11.051. [32] W.E. King, H.D. Barth, V.M. Castillo, G.F. Gallegos, J.W. Gibbs, D.E. Hahn, C. Kamath, [6] W. Gao, Y. Zhang, D. Ramanujan, K. Ramani, Y. Chen, C.B. Williams, C.C.L. Wang, Y.C. A.M. Rubenchik, Observation of keyhole-mode laser melting in laser powder-bed fu- Shin, S. Zhang, P.D. Zavattieri, The status, challenges, and future of additive sion additive manufacturing, J. Mater. Process. Technol. 214 (2014) 2915–2925, manufacturing in engineering, Comput. Aided Des. (2015)https://doi.org/10.1016/ https://doi.org/10.1016/j.jmatprotec.2014.06.005. j.cad.2015.04.001. [33] J. Song, Y. Chew, G. Bi, X. Yao, B. Zhang, J. Bai, S.K. Moon, Numerical and experimen- [7] J.J. Marattukalam, A.K. Singh, S. Datta, M. Das, V.K. Balla, S. Bontha, S.K. Kalpathy, Mi- tal study of laser aided additive manufacturing for melt-pool profile and grain orien- crostructure and corrosion behavior of laser processed NiTi alloy, Mater. Sci. Eng. C. tation analysis, Mater. Des. 137 (2018) 286–297, https://doi.org/10.1016/j.matdes. 57 (2015) 309–313, https://doi.org/10.1016/j.msec.2015.07.067. 2017.10.033. [8] E. Louvis, P. Fox, C.J. Sutcliffe, Selective laser melting of aluminium components, J. [34] Y. Liu, Z. Pang, J. Zhang, Comparative study on the influence of subsequent thermal Mater. Process. Technol. 211 (2011) 275–284, https://doi.org/10.1016/j. cycling on microstructure and mechanical properties of selective laser melted 316L jmatprotec.2010.09.019. stainless steel, Appl. Phys. A Mater. Sci. Process. 123 (2017) 688. J.J. Marattukalam et al. / Materials and Design 193 (2020) 108852 11

[35] A. Basak, S. Das, Epitaxy and microstructure evolution in metal additive manufactur- [55] H. Mecking, U.F. Kocks, C. Hartig, Taylor factors in materials with many deformation ing, Annu. Rev. Mater. Res. 46 (2016) 125–149, https://doi.org/10.1146/annurev- modes, Scr. Mater. 35 (1996) 465–471, https://doi.org/10.1016/1359-6462(96) matsci-070115-031728. 00137-6. [36] M.-S. Pham, B. Dovgyy, P.A. Hooper, C.M. Gourlay, A. Piglione, The role of side- [56] K. Yvell, Experimental Studies of Deformation Structures in Stainless Steels Using branching in microstructure development in laser powder-bed fusion, Nat. EBSD, KTH Royal Institute of Technology, 2018 63 (2018. doi:urn:nbn:se:kth: Commun. 11 (2020) 749, https://doi.org/10.1038/s41467-020-14453-3. diva-227663). [37] X. Zhou, K. Li, D. Zhang, X. Liu, J. Ma, W. Liu, Z. Shen, Textures formed in a CoCrMo [57] N. Hansen, X. Huang, Microstructure and flow stress of polycrystals and single crys- alloy by selective laser melting, J. Alloys Compd. 631 (2015) 153–164, https://doi. tals, Acta Mater. 46 (1998) 1827–1836, https://doi.org/10.1016/S1359-6454(97) org/10.1016/j.jallcom.2015.01.096. 00365-0. [38] D. Dai, D. Gu, Thermal behavior and densification mechanism during selective laser [58] G.L. Hao, Q.P. Xu, H. Wang, X.Y. Li, Effect of pore structure on mechanical properties melting of copper matrix composites: simulation and experiments, Mater. Des. 55 of porous TiAl, Mater. Sci. Technol. (United Kingdom). 32 (2016) 1592–1596, (2014) 482–491, https://doi.org/10.1016/j.matdes.2013.10.006. https://doi.org/10.1080/02670836.2015.1132588. [39] M. Rappaz, S.A. David, J.M. Vitek, L.A. Boatner, Analysis of solidification microstruc- [59] A.E. Wilson-Heid, T.C. Novak, A.M. Beese, Characterization of the effects of internal tures in Fe-Ni-Cr single-crystal welds, Metall. Trans. A. 21 (1990) 1767–1782, pores on tensile properties of additively manufactured austenitic stainless steel https://doi.org/10.1007/BF02672593. 316L, Exp. Mech. (2018)https://doi.org/10.1007/s11340-018-00465-0. [40] S.-H. Sun, K. Hagihara, T. Nakano, Effect of scanning strategy on texture formation in [60] B. Gorny, T. Niendorf, J. Lackmann, M. Thoene, T. Troester, H.J. Maier, In situ charac- Ni-25 at.%Mo alloys fabricated by selective laser melting, Mater. Des. 140 (2018) terization of the deformation and failure behavior of non-stochastic porous struc- – 307 316, https://doi.org/10.1016/J.MATDES.2017.11.060. tures processed by selective laser melting, Mater. Sci. Eng. A 528 (2011) [41] Nancy Yang, J. Yee, B. Zheng, K. Gaiser, T. Reynolds, L. Clemon, W.Y. Lu, J.M. 7962–7967, https://doi.org/10.1016/j.msea.2011.07.026. Schoenung, E.J. Lavernia, Process-structure-property relationships for 316L stainless [61] S.M. Yusuf, Y. Chen, R. Boardman, S. Yang, N. Gao, Investigation on porosity and mi- steel fabricated by additive manufacturing and its implication for component engi- crohardness of 316L stainless steel fabricated by selective laser melting, Metals – neering, J. Therm. Spray Technol. 26 (2016) 610 626, https://doi.org/10.1007/ (Basel) 7 (2017) 64. s11666-016-0480-y. [62] Y. Liu, Y. Yang, S. Mai, D. Wang, C. Song, Investigation into spatter behavior during [42] D.G. McCartney, J.D. Hunt, Measurements of cell and primary dendrite arm spacings selective laser melting of AISI 316L stainless steel powder, Mater. Des. 87 (2015) fi – in directionally solidi ed aluminium alloys, Acta Metall. 29 (1981) 1851 1863, 797–806, https://doi.org/10.1016/j.matdes.2015.08.086. https://doi.org/10.1016/0001-6160(81)90111-5. [63] A.R. Nassar, M.A. Gundermann, E.W. Reutzel, P. Guerrier, M.H. Krane, M.J. Weldon, [43] K. Saeidi, X. Gao, Y. Zhong, Z.J. Shen, Hardened austenite steel with columnar sub- Formation processes for large ejecta and interactions with melt pool formation in grain structure formed by laser melting, Mater. Sci. Eng. A 625 (2015) 221–229, powder bed fusion additive manufacturing, Sci. Rep. 9 (2019) 5038, https://doi. https://doi.org/10.1016/j.msea.2014.12.018. org/10.1038/s41598-019-41415-7. [44] J. Ye, Y.M. Wang, R.T. Ott, Z. Zeng, J.T. McKeown, P.J. Depond, M.K. Santala, T.T. [64] M. Dal, V. Gunenthiram, R. Fabbro, I. Koutiri, F. Coste, P. Peyre, M. Schneider, Exper- Roehling, T. Zhu, N.P. Calta, M.J. Matthews, Y. Zhang, Z. Li, W. Chen, A.V. Hamza, T. imental analysis of spatter generation and melt-pool behavior during the powder Voisin, Additively manufactured hierarchical stainless steels with high strength bed laser beam melting process, J. Mater. Process. Technol. 251 (2017) 376–386, and ductility, Nat. Mater. 17 (2017) 63–71, https://doi.org/10.1038/nmat5021. https://doi.org/10.1016/j.jmatprotec.2017.08.012. [45] Y.J. Yin, J.Q. Sun, J. Guo, X.F. Kan, D.C. Yang, Mechanism of high yield strength and [65] A.B. Anwar, Q. Pham, Effect of inert gas flow velocity and unidirectional scanning on yield ratio of 316 L stainless steel by additive manufacturing, Mater. Sci. Eng. A the formation and accumulation of spattered powder during selective laser melting, 744 (2019) 773–777, https://doi.org/10.1016/j.msea.2018.12.092. Proc. 2nd Int. Conf. Prog. Addit. Manuf. (pro-AM 2016), 2016. [46] Z. Sun, X. Tan, S.B. Tor, C.K. Chua, Simultaneously enhanced strength and ductility for [66] A. Bin Anwar, I.H. Ibrahim, Q.-C. Pham, Spatter transport by inert gas flow in selec- 3D-printed stainless steel 316L by selective laser melting, NPG Asia Mater. 10 tive laser melting: a simulation study, Powder Technol. 352 (2019) 103–116, (2018) 127–136, https://doi.org/10.1038/s41427-018-0018-5. https://doi.org/10.1016/j.powtec.2019.04.044. [47] L. Liu, Q. Ding, Y. Zhong, J. Zou, J. Wu, Y.L. Chiu, J. Li, Z. Zhang, Q. Yu, Z. Shen, Dislo- cation network in additive manufactured steel breaks strength–ductility trade-off, [67] T. Heeling, M. Gerstgrasser, K. Wegener, Investigation of Selective Laser Melting Mater. Today 21 (2018) 354–361, https://doi.org/10.1016/j.mattod.2017.11.004. Spatter Characteristics for Single-and Multi-Beam Strategies Using High Speed Im- [48] M. Shamsujjoha, S.R. Agnew, J.M. Fitz-Gerald, W.R. Moore, T.A. Newman, High aging, in: n.d. strength and ductility of additively manufactured 316L stainless steel explained, [68] D. Gu, Y. Shen, Balling phenomena in direct laser sintering of stainless steel powder: – Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 49 (2018) 3011–3027, https://doi. metallurgical mechanisms and control methods, Mater. Des. 30 (2009) 2903 2910. org/10.1007/s11661-018-4607-2. [69] X. Gong, T. Anderson, K. Chou, Review on powder-based electron beam additive [49] Y. Zhong, L. Liu, S. Wikman, D. Cui, Z. Shen, Intragranular cellular segregation net- manufacturing technology, in: ASME/ISCIE 2012 Int. Symp. Flex. Autom., American – work structure strengthening 316L stainless steel prepared by selective laser melt- Society of Mechanical Engineers Digital Collection, n.d.: pp. 507 515. ing, J. Nucl. Mater. 470 (2016) 170–178, https://doi.org/10.1016/j.jnucmat.2015.12. [70] S. Sinha, J.A. Szpunar, N.A.P. Kiran Kumar, N.P. Gurao, Tensile deformation of 316L 034. austenitic stainless steel using in-situ electron backscatter diffraction and crystal [50] X. Wang, J.A. Muñiz-Lerma, O. Sánchez-Mata, M. Attarian Shandiz, M. Brochu, Mi- plasticity simulations, Mater. Sci. Eng. A 637 (2015) 48–55, https://doi.org/10. crostructure and mechanical properties of stainless steel 316L vertical struts 1016/j.msea.2015.04.005. manufactured by laser powder bed fusion process, Mater. Sci. Eng. A 736 (2018) [71] X. Wang, J.A. Muñiz-Lerma, O. Sanchez-Mata, M. Attarian Shandiz, N. Brodusch, R. 27–40, https://doi.org/10.1016/j.msea.2018.08.069. Gauvin, M. Brochu, Characterization of single crystalline austenitic stainless steel [51] X. Zhang, N. Hansen, Y. Gao, X. Huang, Hall-Petch and dislocation strengthening in thin struts processed by laser powder bed fusion, Scr. Mater. 163 (2019) 51–56, graded nanostructured steel, Acta Mater. 60 (2012) 5933–5943, https://doi.org/10. https://doi.org/10.1016/j.scriptamat.2018.12.032. 1016/j.actamat.2012.07.037. [72] A. Barabi, A. Zarei-Hanzaki, H. Abedi, A. Anoushe, J. Cho, The correlation of macro- [52] N. Hansen, X. Huang, G. Winther, Grain orientation, deformation microstructure and structure, microstructure, and texture with room temperature mechanical proper- flow stress, Mater. Sci. Eng. A 494 (2008) 61–67, https://doi.org/10.1016/j.msea. ties of a twinning-induced plasticity automotive steel after friction stir spot 2007.10.086. welding/processing, Steel Res. Int. 89 (2018), 1800245. . [53] F.R.N. Nabarro, Z.S. Basinski, D.B. Holt, The plasticity of pure single crystals, Adv. [73] A. Barabi, A. Zarei-hanzaki, M. Mosayebi, A. Mohamadizadeh, A.S. Anoushe, Novel Phys. 13 (1964) 193–323. analytical approach for evaluating the mechanical properties of friction stir spot [54] R. Neuhaus, C. Schwink, On the flow stress of [100]-and [111]-oriented Cu-Mn sin- joints through constitutive modeling, Eng. Fract. Mech. 216 (2019) 106522, gle crystals: a transmission electron microscopy study, Philos. Mag. A. 65 (1992) https://doi.org/10.1016/j.engfracmech.2019.106522. 1463–1484.